What they do when in doubt: a study of inductive biases in seq2seq learners

arXiv2006.14953MaRDI QIDQ6343814

Author name not available (Why is that?)

Publication date: 26 June 2020

Abstract: Sequence-to-sequence (seq2seq) learners are widely used, but we still have only limited knowledge about what inductive biases shape the way they generalize. We address that by investigating how popular seq2seq learners generalize in tasks that have high ambiguity in the training data. We use SCAN and three new tasks to study learners' preferences for memorization, arithmetic, hierarchical, and compositional reasoning. Further, we connect to Solomonoff's theory of induction and propose to use description length as a principled and sensitive measure of inductive biases. In our experimental study, we find that LSTM-based learners can learn to perform counting, addition, and multiplication by a constant from a single training example. Furthermore, Transformer and LSTM-based learners show a bias toward the hierarchical induction over the linear one, while CNN-based learners prefer the opposite. On the SCAN dataset, we find that CNN-based, and, to a lesser degree, Transformer- and LSTM-based learners have a preference for compositional generalization over memorization. Finally, across all our experiments, description length proved to be a sensitive measure of inductive biases.

Has companion code repository: https://github.com/facebookresearch/FIND

This page was built for publication: What they do when in doubt: a study of inductive biases in seq2seq learners

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6343814)